1. Field of the Invention
The field of the present invention pertains to communications and, more particularly, to an air interface structure and protocol suitable for use in a cellular communication environment.
2. Description of Related Art
A growing demand for flexible, mobile communication has led to development of a variety of techniques for allocating available communication bandwidth among a steadily increasing number of users of cellular services. Two conventional techniques for allocating communication bandwidth between a cellular base station and a set of cellular user stations (also called "mobile stations") are frequency division duplex (FDD) and time division duplex (TDD).
As used herein, FDD refers to a technique for establishing full duplex communications having both forward and reverse links separated in frequency, and TDD refers to a technique for establishing full duplex communications having both forward and reverse links occurring on the same frequency but separated in time to avoid collisions. Other techniques for communication are time division multiple access (TDMA), wherein transmissions by a plurality of users are separated in time to avoid conflicts, frequency division multiple access (FDMA) wherein transmissions by a plurality of users are separated in frequency to avoid conflicts, and time division multiplex (TDM), wherein multiple data streams are time multiplexed together over a single carrier. Various combinations of FDD, TDD, FDLA, and TDMA may also be utilized.
In a particular FDD technique, a base station is allocated a set of frequencies over which it may transmit, using a different frequency slot for each user station, and each user station is allocated a different frequency over which it may transmit to the base station. For each new user in contact with a base station, a new pair of frequencies is required to support the communication link between the base station and the new user station. The number of users that can be supported by a single base station is therefore limited by the number of available frequency slots.
In a particular TDD technique, the same frequency is used for all user stations in communication with a particular base station. Interference between user stations is avoided by requiring that user stations transmit at different times from one another and from the base station. This is accomplished by dividing a time period into a plurality of time frames, and each time frame into a plurality of time slots. Typically, the base station communicates with only one user station during a time slot, and communicates with all the user stations sequentially during different time slots over a single time frame. Thus, the base station communicates with a particular user station once during each time frame.
In one version of the described system, the base station is allocated a first portion of each time slot during which the base station transmits to a particular user station, and the user station is allocated a second portion of the time slot during which the user station responds to the base station. Thus, the base station may transmit to a first user station, await a response, and, after receiving a response from the first user station, transmit to a second user station, and so on, until the base station has communicated with all user stations sequentially over a particular time frame.
Time division duplex has an advantage over FDD and FDMA of requiring use of only a single frequency bandwidth. However, a drawback of many conventional TDD or TDMA systems is that their efficiency suffers as cell size increases. The reduction in efficiency stems from the relatively unpredictable nature of propagation delay times of transmissions from the base station over air channels to the user stations, and from the user stations over air channels back to the base station. Because user stations are often mobile and can move anywhere within the radius of the cell covered by a base station, the base station generally does not know in advance how long the propagation delay will be for communicating with a particular user station. In order to plan for the worst case, conventional TDD systems typically provide a round-trip guard time to ensure that communication will be completed with the first user station before initiating communication with the second user station. Because the round-trip guard time is present in each time slot regardless of how near or far a user station is, the required round-trip guard time can add substantial overhead, particularly in large cells. The extra overnead limits the number of users, and hence the efficiency, of TDD systems.
FIG. 1 is an illustration of the basic round trip timing for a TDD system from a base station perspective. A polling loop 101, or time frame, for a base station is divided into a plurality of time slots 103. Each time slot 103 is used for communication from the base station to a particular user station. Thus, each time slot comprises a base transmission 105, a user transmission 107, and a delay period 106 during which the base transmission 105 propagates to the user station, the user station processes and generates a responsive user transmission 107, and the user transmission 107 propagates to the base station.
If the user station is located right next to the base station, then the, base station can expect to hear from the user station immediately after finishing its transmission and switching to a receive mode. As the distance between the user station and the base station grows, the time spent by the base station waiting for a response grows as well. The base station will not hear from the user station immediately but will have to wait for signals to propagate to the user station and back.
As shown in FIG. 1, in a first time slot 110 the user transmission 107 arrives at the base station at a time approximately equidistant between the end of the base transmission 105 and the start of the user transmission 107, indicating that the user station is about half a cell radius from the base station. In a second time slot 111, the user transmission 107 appears very close to the end of the base transmission 105, indicating that the user station is very close to the base station. In a third time slot 112, the user transmission 107 appears at the very end of the time slot 112, indicating that the user station is near or at the cell boundary. Because the third time slot 112 corresponds to a user station at the maximum communication distance for a particular base station, the delay 106 shown in the third time slot 112 represents the maximum round-trip propagation time and, hence, the maximum round-trip guard time.
In addition to propagation delay times, there also may be delays in switching between receive and transmit mode in the user station, base station, or both, which are not depicted in FIG. 1 for simplicity. Typical transmit/receive switching times are about two microseconds, but additional allocations may be made to account for channel ringing effects associated with multipath.
As cell size increases, TDD guard time must increase to account for longer propagation times. In such a case, guard time consumes an increasingly large portion of the available time slot, particularly for shorter round trip frame durations. The percentage increase in time spent for overhead is due to the fact that TDD guard time is a fixed length, determined by cell radius, while the actual round trip frame duration varies according to the distance of the user station. Consequently, as cells get larger, an increasing amount of time is spent on overhead in the form of guard times rather than actual information transfer between user stations and the base station.
One conventional TDD system is the Digital European Cordless Telecommunications (DECT) system developed by the European Telecommunications Standards Institute (ETSI). In the DECT system, a base station transmits a long burst of data segmented into time slots, with each time slot having data associated with a particular user station. After a guard time, user stations respond in a designated group of consecutive time slots, in the same order as the base station sent data to the user stations.
Another system in current use is the Global System for Mobile communications ("GSM"). FIG. 4 illustrates a timing pattern according to certain existing GSM standards. According to these standards, communication between a base station and user stations is divided into eight burst periods 402. Up to eight different user stations can communicate with a base station, one in each burst period 402.
GSM standards require two separate frequency bands. The base station transmits over a first frequency F.sub.A, while the user stations transmit over a second frequency F.sub.B. After a user station receives a base transmission 405 on the first frequency F.sub.A during a particular burst period 402, the user station shifts in frequency by 45 MHz to the second frequency F.sub.B and transmits a user transmission 406 in response to the base transmission 405 approximately three burst periods 402 later. The three burst period delay is assumed to be large enough to account for propagation time between the base station and the user station.
It is important in the GSM system that the user transmissions 406 received at the base station fit into the appropriate burst periods 402. Otherwise, the user transmissions 406 from user stations using adjacent burst periods 402 could overlap, resulting in poor transmission quality or even loss of communication due to interference between user stations. Accordingly, each burst period 402 is surrounded by a guard times 407 to account for uncertain signal propagation delays between the base station and the user station. By comparing the time of the signal actually received from the user station 302 to the expected receive time, the base station may command the user station to advance or retard its transmission timing in order to fall within the proper burst period 402, a feature known as adaptive frame alignment. A specification relating to adaptive frame alignment for the GSM system is TS GSM 05.10.
A drawback of the described GSM system is that it requires two separate frequency bands. It also has a relatively rigid structure, which may limit its flexibility or adaptability to certain cellular environments.
Another system in presence use is known as Wide Area Coverage System (WACS), a narrowband system employing aspects of both FDMA and TDMA. Under WACS, as in GSM, two distinct frequency bands are used. One frequency band is used for user station transmissions, and the other frequency band is used for base station transmissions. The user station transmissions are offset by one-half of a time slot from the corresponding base station transmissions, in order to allow for propagation time between the base station and the user station. Standard WACS does not support spread spectrum communication (a known type of communication wherein the bandwidth of the transmitted signal exceeds the bandwidth or the data to be transmitted) , and has an overall structure that may be characterized as relatively rigid.
In a number of systems, the channel structure is such that a user station may have to transmit a response to a base station while receiving information on another channel. The capability for simultaneous transmission and reception generally requires the use of a diplexer, which is a relatively expensive component for a mobile handset.
It would be advantageous to provide a flexible system having the benefits of time division duplex communication, particularly in large cells, but without having an overhead of a full round-trip guard time in every time slot. It would further be advantageous to provide such a system requiring only a single frequency band for communication. It would further be advantageous to provide a TDMA or combination TDMA/FDMA system wherein user stations are not required to be fitted with a diplexer. It would further be advantageous to provide a time frame structure readily adaptable to single or multiple frequency bands, and for use in either a variety of communication environments.